Mortgage lender Newrez is expanding its use of AI to address the pain points in underwriting processes, marking the next step of its partnership with technology provider HomeVision.  

Newrez began using HomeVision’s AI tool designed for collateral underwriting in September and has already seen efficiency double, according to a Jan. 7 Newrez release. Newrez invested an undisclosed amount in HomeVision this month to develop an enhanced solution that also synthesizes data for income, borrower assets and credit, with plans to deploy the new tech later this year.  

The lender is targeting similar efficiency gains as the platform, which uses machine learning to make decisions, automates these additional categories, Chief Information Officer Brian Woodring told FinAi News.  

“I want a mortgage [loan] to become something that’s done in weeks, not a month or more, and I think that’s very achievable,” he said.  

Fort Washington, Pa.-based Newrez evaluates these factors to measure efficiency, Woodring said: 

• Speed of loan originations and approvals; 
• Number of human touches or oversights;  
• Overall time cycle of mortgage fulfillment; and 
• Scalability. 

With the expanded platform, Newrez also aims to lower origination costs, enhance accuracy and create a more seamless customer experience, according to the release. 

Mortgage lending pain points 

The global market for AI-powered mortgage underwriting is projected to grow 22.6% annually to $13.5 billion in 2033, according to market research and consulting firm DataIntelo.  

AI is suited for mortgage underwriting due to the amount of data, Woodring said.  

“Underwriting a home loan is one of the most complex financial transactions in consumer finance,” he said. “There are hundreds and hundreds of data points. … What HomeVision really built is a process that takes the biggest piece of work there, which is the underwrite, and really automates a lot of that.” 

Scalability 

Automating underwriting lets mortgage professionals focus more on customer service, “making sure the right decision is being made and the risk is being managed correctly,” Woodring said. 

Further, scalability is crucial in mortgage lending because of the industry’s cyclical nature, Woodring said.  

“It’s very difficult to react to market conditions and scale up and scale down,” he said.  

To overcome this hurdle, Newrez and HomeVision apply the “elasticity” of cloud technology to the platform, Woodring said.  

Refreshing data as part of “reinforcement learning” is also pivotal as regulations, investor guidelines and other industry shifts frequently occur, he said.  

“The models are optimized to provide the most accurate analysis possible, and those always have to be maintained,” he said. 

Better borrower experience 

AI can improve the borrower experience by accelerating the mortgage lending process, eliminating much of the angst that homebuyers commonly experience while waiting on a mortgage approval, Woodring said. 

In addition, AI can improve interactions between borrowers and lenders by minimizing tedious follow-up questions and documents, he said. 

For example, “if we were to find out in the middle of an underwrite that you had been divorced and we didn’t know that before, well, suddenly we need your divorce decree,” he said. “We need all of your divorce information because that may have a debt or a liability.” 

Newrez and HomeVision considered these types of friction points when developing the expanded platform, Woodring said.  

Register here for the inaugural FinAi Banking Summit, taking place March 2-3 in Denver. View the full event agenda here.